Monetary crime and fraud prevention specialist Feedzai unveiled its RiskFM (Danger Foundational Mannequin) answer this week.
RiskFM covers a broad vary of economic information to supply danger decisioning throughout fraud detection, anti-money laundering, and different monetary crime.
Headquartered in New York and based in 2008, Feedzai made its Finovate debut at FinovateEurope 2014 in London.
Monetary crime prevention innovator Feedzai launched its RiskFM (Danger Foundational Mannequin) answer this week. The brand new providing leverages a Tabular Basis Mannequin that’s purpose-built for monetary information and danger decisioning, altering the way in which that monetary crime is detected and prevented.
Spanning throughout fraud detection, anti-money laundering (AML), and different monetary crime-related danger choices, RiskFM is skilled on a broad, deep, world dataset masking onboarding, digital exercise, funds, fund transfers, and AML workflows to allow establishments to determine, forestall, and adapt to monetary crime rapidly and precisely.
The answer is designed to deal with a few of the particular challenges of coping with transactional information. Of their assertion asserting the brand new providing, Feedzai in contrast this problem with giant language fashions (LLMs) and their potential to take care of domains corresponding to language, audio, and video. These domains, the corporate famous, all have finite grammar and a sure linear causality. Against this, monetary transactions are far much less predictive, largely as a result of the patron conduct behind these transactions, from cost sorts to fraud modalities, can and does change—steadily.
“Subsequent transactions are far much less predictable than the subsequent phrase in a sentence,” Feedzai Chief Science Officer Pedro Bizarro mentioned. “Client spending habits, cost sorts, and fraud modes change constantly. Extra importantly, monetary danger is an adversarial area; fraudsters actively adapt to evade detection in actual time.”
The flexibility to function throughout a number of establishments and geographies on the identical time is one key characteristic of RiskFM, and when used to energy a custom-made mannequin for a single buyer, RiskFM matches the efficiency of high-tuned, supervised fashions whereas avoiding time-consuming, handbook characteristic engineering. RiskFM outperformed conventional fashions based mostly on Gradient Boosting and Deep Studying methods, and is constructed for the total vary of economic crime prevention, from mule account detection to anti-money laundering. The corporate refers back to the expertise because the “foundational AI layer for monetary danger,” making certain establishments have an clever, scalable answer that grows as they do.
“RiskFM proves our multi-year funding in basis fashions is paying off,” Feedzai Chief Product Officer Pedro Barata mentioned. “We’re not simply a part of the dialog; we’re defining the way it applies to the complexities of world monetary crime prevention.”
Feedzai made its Finovate debut at FinovateEurope 2014. Headquartered in New York and based in 2008, Feedzai right now presents an AI-native monetary crime prevention platform that helps banks, cost networks, acquirers, and different monetary companies suppliers detect and stop monetary crime, fraud, and cash laundering in actual time. The corporate’s platform serves multiple billion shoppers, processes 90 billion occasions, and secures $9 trillion in cost quantity yearly.
Within the wake of its RiskFM announcement, the corporate since reported that it has been named to Quick Firm’s World’s Most Revolutionary Corporations 2026 roster. “We at Feedzai are honored by this prestigious recognition of our innovation and analysis in trusted AI to construct a world of safer cash,” Feedzai Co-Founder and CEO Nuno Sebastiao mentioned.
Picture by Tima Miroshnichenko
Views: 55
